The need and demand for an accurate, non-invasive method for determining analyte concentrations in human tissue is well documented. Barnes et al. (U.S. Pat. No. 5,379,764), for example, disclose the necessity for diabetics to frequently monitor glucose levels in their blood. It is further recognized that the more frequent the analysis and subsequent medication, the less likely there will be large swings in glucose levels. These large swings are associated with symptoms and complications of the disease, whose long term effects can include heart disease, arteriosclerosis, blindness, stroke, hypertension, kidney failure, and premature death. As described below, systems have been proposed for the non-invasive measurement of glucose in blood. However, despite these efforts, a lancet cut into the finger is still necessary for all presently commercially available forms of home glucose monitoring. This is believed so compromising to the diabetic patient that the most effective use of any form of diabetic management is rarely achieved.
The various proposed non-invasive methods for determining blood glucose level, discussed individually below, generally utilize quantitative infrared spectroscopy as a theoretical basis for analysis. Infrared spectroscopy measures the electromagnetic radiation (0.7-25 .mu.m) a substance absorbs at various wavelengths. Atoms do not maintain fixed positions with respect to each other, but vibrate back and forth about an average distance. Absorption of light at the appropriate energy causes the atoms to become excited to a higher vibration level. The excitation of the atoms to an excited state occurs only at certain discrete energy levels, which are characteristic for that particular molecule. The most primary vibrational states occur in the mid-infrared frequency region (i.e., 2.5-25 .mu.m). However, non-invasive analyte determination in blood in this region is problematic, if not impossible, due to the absorption of the light by water. The problem is overcome through the use of shorter wavelengths of light which are not as attenuated by water. Overtones of the primary vibrational states exist at shorter wavelengths and enable quantitative determinations at these wavelengths.
It is known that glucose absorbs at multiple frequencies in both the mid- and near-infrared range. There are, however, other infrared active analytes in the blood which also absorb at similar frequencies. Due to the overlapping nature of these absorption bands, no single or specific frequency can be used for reliable non-invasive glucose measurement. Analysis of spectral data for glucose measurement thus requires evaluation of many spectral intensities over a wide spectral range to achieve the sensitivity, precision, accuracy, and reliability necessary for quantitative determination. In addition to overlapping absorption bands, measurement of glucose is further complicated by the fact that glucose is a minor component by weight in blood, and that the resulting spectral data may exhibit a non-linear response due to both the properties of the substance being examined and/or inherent non-linearities in optical instrumentation.
Robinson et al. (U.S. Pat. No. 4,975,581) disclose a method and apparatus for measuring a characteristic of unknown value in a biological sample using infrared spectroscopy in conjunction with a multivariate model that is empirically derived from a set of spectra of biological samples of known characteristic values. The above-mentioned characteristic is generally the concentration of an analyte, such as glucose, but also may be any chemical or physical property of the sample.
The method of Robinson et al. involves a two-step process that includes both calibration and prediction steps. In the calibration step, the infrared light is coupled to calibration samples of known characteristic values so that there is differential attenuation of at least several wavelengths of the infrared radiation as a function of the various components and analyte comprising the sample with known characteristic value. The infrared light is coupled to the sample by passing the light through the sample or by reflecting the light from the sample. Absorption of the infrared light by the sample causes intensity variations of the light that are a function of the wavelength of the light. The resulting intensity variations at the at least several wavelengths are measured for the set of calibration samples of known characteristic values. Original or transformed intensity variations are then empirically related to the known characteristic of the calibration samples using a multivariate algorithm to obtain a multivariate calibration model.
In the prediction step, the infrared light is coupled to a sample of unknown characteristic value, and the calibration model is applied to the original or transformed intensity variations of the appropriate wavelengths of light measured from this unknown sample. The result of the prediction step is the estimated value of the characteristic in the unknown sample. The disclosure of Robinson et al. is incorporated herein by reference.
Dahne et al. (U.S. Pat. No. 4,655,225) further disclose a method utilizing near infrared spectroscopy for non-invasively transmitting optical energy in the near infrared spectrum through a finger or earlobe of a subject. Dahne also disclose measuring reflected light energy to determine analyte concentration. The reflected light energy is further stated as comprised of light reflected from the surface of the sample and light reflected from deep within the tissue. It is the near infrared energy diffusely reflected from deep within the tissues that Dahne disclose as containing analyte information, while surface reflected light energy gives no analyte information and interferes with interpreting or measuring light reflected from deep in the tissue. The present invention is directed to an apparatus for improved measurement of diffusely reflected light, while eliminating the effects of surface reflected light and other light not reflected from deep within the tissue.
Reflectance spectroscopy is known in other non-medical applications. In general, such spectroscopy is concerned with identification of the chemical structure of the sample through the use of reflected information. Diffuse reflectance spectroscopy is also generally known, and is widely used in the visible and near-infrared regions of the light spectrum to study materials such as grains and other food products.
In broad terms, diffuse reflectance spectroscopy utilizes the fact that the sample materials will tend to scatter light in a more or less random fashion. A fraction of the light will eventually be scattered back from the sample and collected by a detector to provide a quantitative or qualitative representation of the sample.
In infrared spectroscopy it is often desirable to use the mid-infrared region of the spectrum. The fundamental vibrational absorptions described earlier are strongest here, in the fundamental region. The goal of infrared spectroscopy sampling is often to prepare a sample so that it may be analyzed with this mid-infrared light. Reflectance spectroscopy is one very popular way of making a sample compatible with mid-infrared light. If a sample is too thick to get any light through in transmission, often a result can be obtained by reflectance. Reflectance spectroscopy is complicated however, by the fact that there is more than one optical phenomenon occurring in this mode.
Reflectance of light from a sample can be largely divided into two categories, diffuse reflectance and specular reflectance. The specular reflectance of a sample is the light which does not propagate into the sample, but rather reflects "like a mirror" from the front surface of the sample. This component contains information about the sample at the surface. If the material is homogeneous, this surface reflection can be related to the bulk. While the specular component does not physically appear much like an absorbance spectrum, it can be related to the absorbance spectrum of the bulk material through a transformation called the Kramers-Kronig transformation. Still, most experts agree that the diffuse component is much more useful for sample qualification and quantification than is the specular component. There has been a lot of effort, by the applicants and by others, to enhance the diffuse component, and de-emphasize the specular component and to essentially cause the reflectance spectrum to be more transmission-like.
Generally these efforts fall largely into three categories: optical discrimination against specular, mechanical discrimination, and secondary methods of sample preparation designed to minimize specular. A fourth, non-independent approach is to move away from the mid-infrared region in order to relax the sample preparation requirements. By moving to the near-infrared or visible region of the spectrum, the vibrational spectroscopy becomes more blunt and imprecise, but often this can be made up for by the improvements observed in the quality and signal-to-noise ratio of the data obtained because of improved sampling ability, more appropriate path length, and better discrimination against specular reflectance. This approach is especially useful when quantitative information is desired.
Most experts would agree that the diffuse component is desirable, and even essential, if the sample material is layered or non-homogeneous. The specular component will largely contain information about the surface of the sample and not about the bulk. Nevertheless, U.S. Pat. No. 5,015,100, issued May 14, 1991 to Walter M. Doyle, describes an example of the specular approach. The specular component of the light is significantly wavelength dependent, and contains information about the complex refractive index of the material under test. This complex refractive index contains an imaginary term which relates to the absorption coefficient of the material.
Doyle indicates that the potential utility of specular reflectance spectroscopy is well-known to those of skill in the art and points out that mathematical expressions, namely the Kramers-Kronig relation, can be used to convert measured reflectance spectra into absorbance spectra. These calculated spectra are then useful for identifying samples by comparison with existing libraries of absorbance spectra. However, the work of the prior art has not been used for quantitative measurements such as the composition analysis of tissue fluids. In fact, it would perform poorly for this purpose, since there is little tissue fluid information at the surface of the skin. The diffuse component must be used.
Paper No. 424, presented at the 16th Annual FACSS Conference in October, 1989, by Doyle and McIntosh, concluded that the Kramers-Kronig relations could not be used to obtain accurate absorbance spectra from reflectance data unless the equations used were modified to take into consideration polarization and angle of incidence, or unless the experimental apparatus provided radiation which approximated the conditions at normal incidence.
The Doyle patent reference describes the use of apparatus in a specular reflectance system in which the analytical radiation reflected by the sample approximates the conditions existing at normal incidence, and proposed a solution by ensuring essentially equal contributions from rays polarized parallel to the plane of incidence and from rays polarized perpendicular to the plane of incidence. Doyle teaches that a semi-transparent beamsplitter used in such an apparatus would achieve the desired polarization balance, but would sacrifice radiation efficiency because of losses in pre-sample transmission, post-sample reflection, and absorbance loss in the beamsplitter. The Doyle reference then described a system of improved radiation efficiency utilizing a split field beamsplitter having a surface area divided into an uneven plurality of reflecting blades and open transmitting areas.
U.S. Pat. No. 4,852,955 also issued to Doyle, describes a system which obviates the problem of limited beamsplitter efficiency by using a 100% reflecting mirror intercepting half of the system aperture, and arranging for the illuminating and outgoing beams to use opposite halves of the aperture. However, the use of the split field beamsplitter of this reference involves a distribution of incident radiation which is asymmetrical with respect to an axis normal to the sample surface. As a result, there is no assurance that the p and s polarization states will be balanced when the suggested beamsplitter is in use.
The limitations of Doyle's prior art are clear. Specular reflectance is only useful when the bulk material is adequately represented by surface composition. When this is not the case, such as when performing non-invasive blood analyte measurements, this methodology will give a spurious result.
Optical means have also been used to separate diffuse and specular components. A recent example is described by Ralf Marbach in his PhD. thesis entitled "Messverfahren zur IR-spektroskopishen Blutglucose bestimmung" (English translation: "Measurement Techniques for IR Spectroscopic Blood Glucose Determination"), and published in Duesseldorf in 1993. Marbach employs an optical discrimination system quite similar in principle to that used by Harrick Scientific Corp. in the Praying Mantis diffuse reflectance instrument first introduced in 1980. The concept here is that the specular light reflects from a sample with an angle equal and opposite to the angle of incidence to the surface normal. Using this fact, it is a simple matter to collect light only outside the input collection angle. Marbach and Harrick then limit the input angle to a small range, so that a larger range of output angles may be used for collection.
Note that there is a limited region of space over which light can be launched into and collected from a sample. In terms of solid angle, for a planar surface sample, this working volume can be stated to be 2.pi. steradians in solid angle. In the Harrick device, a small and equal solid angle is subtended by the input and the output optics. Less than 1/2.pi. steradians is subtended by either the input or the output optic. This leads to an efficiency of less than 50% of the available solid angle. Another critical factor in collecting diffusely reflected light is the directionality of the collected light. Many samples, including the tissue samples required for non-invasive measurements are quite forward scattering. That is to say that a scattered photon will change only a small angle in direction after a scattering event. The Harrick device requires a photon to deviate through a large angle before it can be collected by the output optics. This poor performance in the presence of sample anisotropy and the relatively low efficiency are severe problems with the Harrick device.
The Marbach device improves on the Harrick device in a number of ways. First, the total volume available for input and collection of light approaches 2.pi. steradians which is the theoretical limit. This is accomplished by allowing 360.degree. azimuthal angular subtense for both the input and output light. Second, the forward directionality of scatter is taken into account. Rays which deviate only a few degrees in angle can be collected. The downfall of this approach is that the input and output optical systems are completely unmatched in terms of magnification. Any diffuse reflectance system must work in concert with the source and the detector of the system.
Since detectors in the near-infrared region of the spectrum get noisier when they get bigger, it should be a goal to make the detector as small as possible. A bright compact source is also advantageous. In the Marbach system, the image of the source is very much magnified relative to the image of the detector in the sample plane. This means that the source energy density which can be imaged onto the detector is limited. In addition, the collected energy from the sample is demagnified as it travels to the detector. Again, energy efficiency is compromised. An ideal situation would leave the input and output magnifications equal.
Another important limitation of the Marbach design relates to the choice of angles for input and output. Real optical systems are good at imaging with large f/numbers. Small f/number systems, especially with large field stop diameters, tend to image poorly. Marbach notes this fact in his thesis. In his design, the prime, large f/number, near-normal space is all reserved for input light, and the non-ideal near-grazing light is used for output. It is quite conceivable that the device would work better if used "backwards" from the mode employed by Marbach, where the source site and the detector site would be switched. The device described in this application provides an even better solution.
Another method of eliminating specular contribution to a diffuse reflectance spectrum is to modify the sample itself to reduce its propensity to reflect specularly. One way to accomplish this is to dilute a powdered sample in a non-absorbing matrix material with a low refractive index. The low index matrix will have a low amount of specular component and will mitigate the specular problem. Unfortunately, the goal of non-invasive analysis does not allow for modification of the sample, and so in the field of use described here, these dilution methods are not an option.
Finally, an apparatus for mechanically discriminating against specular reflectance is shown in U.S. Pat. No. 4,661,706, issued Apr. 28, 1987, to Robert G. Messerschmidt and Donald W. Sting. Messerschmidt et al. demonstrate that the specular and the diffuse component of reflected light can be separated mechanically, taking advantage of the fact that the specular component emanates from the surface of the sample. A blade-like device, or blocker, "skims" the specular light before it can impinge on the detector.
Messerschmidt et al. teach that a "thin" blocker is essential to maximizing the efficiency of the system, and minimizing the distortion of the output spectrum. More particularly, Messerschmidt et al. state that to obtain the maximum efficiency and the closest approximation to the Kubelka-Munk relationship, a thin blocker device should be used having an edge that is a fraction of the optical depth of the sample. A thicker blocker, Messerschmidt et al. explain, will remove energy that penetrates only a short distance into the sample before reflecting, and thus may have a catastrophic effect on the efficiency when used with a sample having a shallow optical depth.
Messerschmidt et al. also state that a thick blocker may introduce spectral distortions caused by energy that is once reflected by the sample to the lower surface of the blocker and again reflected from the blocker to the sample before energy escapes from the far side of the blocker. This is problematic, according to Messerschmidt et al., because the energy reflected from the lower surface of the blocker will acquire the reflectance spectral features of the blocker itself and thus distort the output spectrum.
Applicants have discovered that the "thin" blocker approach of Messerschmidt et al. suffers from a number of limitations, some of which are discussed below. First, the "thin" blocker approach does not provide any discrimination between the diffusely reflected energy that is reflected from various depths within the sample. This limitation is of particular importance when the sample is layered or otherwise non-homogeneous, and only a selected set of the layers contain the desired information. Second, the "thin" blocker of Messerschmidt et al. may not perfectly conform to a rough surface of a sample. This can cause locations where the light effectively leaks or pipes under the blocker without interacting with the sample, thereby distorting the resulting output spectrum.